The present application generally relates to integrated circuit (IC) chips, and more particularly to a method of writing a unique chip identification (ID) in an optically readable format on a top level of an IC chip.
During semiconductor device fabrication a large number of IC chips are concurrently formed on wafer substrates. IC chips are typically much smaller than a single wafer, and consequently, tens of thousands of chips can be fabricated using a single wafer. The actual number of IC chips yielded from a single wafer is a function of the wafer size and the individual chip size. Wafer manufacturers typically mark bare wafers with a unique code or identifier using, e.g., laser impingement. This identification code, typically located on the wafer edge, is unique to each wafer and can be human-readable, machine-readable, or both.
The wafers are then diced into individual IC chips. Once wafers are diced, the IC chips are picked and assembled into chip packages. At this point, it is not easy to determine exactly where on a wafer a particular IC chip originated. Determining the original wafer location of an IC chip is known as chip location identification. There are several reasons why it is desirable to have chip location identification traceability. For example, such information is useful for process learning so that defects can be corrected, product dispositioning for known defects, qualification learning, and engineering evaluation support (e.g., ‘wafer striping’).